yade-dev team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

Do you confirm Vaclav, so I can I commit?

B.

Bruno Chareyre a écrit :
> Corresponding output (after fix). It looks better :
>
> trialFs.norm(): 36.0967 vs. maxFs: 24.6416 Fn: 281.663 ks: 341874 
> tanFric: 0.0874861
> displacementT() before slip: 0.000105585
> displacementT() after slip: 7.2078e-05
> corrected trialFs: 12.4187 -15.884 -14.1663
>
>
>
> Bruno Chareyre a écrit :
> > (discussion moved to yade-dev)
> >
> > Dem3DofGeom_SphereSphere.cpp:76 reads :
> >
> > Vector3r diff=.5*(sqrt(currDistSq)/displacementTMax-1)*(p2-p1);
> >    setTgPlanePts(p1+diff,p2-diff);
> >
> > What about this instead :
> >
> > Vector3r diff=.5*(displacementTMax/sqrt(currDistSq)-1)*(p2-p1);
> >    setTgPlanePts(p1-diff,p2+diff);
> >
> > It gives stable results and no effect of friction. :)
> > Still curious about the previous "weighted average".
> >
> > Bruno
> >
> >
> >
> >
> > Bruno Chareyre a écrit :
> > > The reason is still not clear but I think I located the problem. 
> > > slipToDisplacementTmax is not doing what it should.
> > >
> > > Outputing data from inside the if(plasticity) condition of 
> > > Law2_Dem3DofGeom_FrictPhys_Basic, I find this :
> > >
> > >
> > > trialFs.norm(): 113037 vs. maxFs: 14.9277 Fn: 170.629 ks: 338110 
> > > tanFric: 0.0874861
> > > displacementT() before slip: 0.33432
> > > displacementT() after slip: 0.126106
> > > corrected trialFs: -4.40076 9.24839 10.8599
> > >
> > > The force is divided by ~10,000, as it should, but the displacement 
> > > returned by displacementT is only divided by 3.
> > > Actually, trialFs is always orders of magnitude higher than maxFs, 
> > > even after many iterations in a stable packing. It explains why 
> > > changing friction gives higher forces immediatly.
> > >
> > > I've been looking into relocateContactPoints, could you explain 
> > > this. What is this equation for? It looks like a weird weighted 
> > > average :
> > >
> > > /Vector3r midPt=(effR1/(effR1+effR2))*(p1+p2); // proportionally to 
> > > radii, so that angle would be the same/
> > >
> > > Cheers.
> > >
> > > Bruno
> > >
> > >
> > >
> > >
> > > _______________________________________________
> > > Mailing list: https://launchpad.net/~yade-users
> > > Post to     : yade-users@xxxxxxxxxxxxxxxxxxx
> > > Unsubscribe : https://launchpad.net/~yade-users
> > > More help   : https://help.launchpad.net/ListHelp


--
_______________
Bruno Chareyre
Associate Professor
Grenoble INP
Lab. 3SR
BP 53 - 38041, Grenoble cedex 9 - France
Tél : 33 4 56 52 86 21
Fax : 33 4 76 82 70 43
________________